FM features for automatic forensic speaker recognition.
Tharmarajah ThiruvaranEliathamby AmbikairajahJulien EppsPublished in: INTERSPEECH (2008)
Keyphrases
- speaker recognition
- classification accuracy
- gaussian mixture model
- mel frequency cepstral coefficients
- feature space
- feature vectors
- probabilistic neural network
- feature set
- maximum likelihood
- speaker verification
- extracting features
- extracted features
- vector quantization
- image processing
- low level
- prior knowledge
- multiscale
- bayesian networks